Bio-inspired Double Angstrom-Scale Confinement in Ti-deficient Ti0.87 O2 Nanosheet Membranes for Ultrahigh-performance Osmotic Power Generation.

Osmotic power, a clean energy source, can be harvested from the salinity difference between seawater and river water. However, the output power densities are hampered by the trade-off between ion selectivity and ion permeability. Here we propose an effective strategy of double angstrom-scale confinement (DAC) to design ion-permselective channels with enhanced ion selectivity and permeability simultaneously. The fabricated DAC-Ti0.87 O2 membranes possess both Ti atomic vacancies and an interlayer free spacing of ≈2.2 Å, which not only generates a profitable confinement effect for Na+ ions to enable high ion selectivity but also induces a strong interaction with Na+ ions to benefit high ion permeability. Consequently, when applied to osmotic power generation, the DAC-Ti0.87 O2 membranes achieved an ultrahigh power density of 17.8 W m-2 by mixing 0.5/0.01 M NaCl solution and up to 114.2 W m-2 with a 500-fold salinity gradient, far exceeding all the reported macroscopic-scale membranes. This work highlights the potential of the construction of DAC ion-permselective channels for two-dimensional materials in high-performance nanofluidic energy systems.

PL-Net: progressive learning network for medical image segmentation.

In recent years, deep convolutional neural network-based segmentation methods have achieved state-of-the-art performance for many medical analysis tasks. However, most of these approaches rely on optimizing the U-Net structure or adding new functional modules, which overlooks the complementation and fusion of coarse-grained and fine-grained semantic information. To address these issues, we propose a 2D medical image segmentation framework called Progressive Learning Network (PL-Net), which comprises Internal Progressive Learning (IPL) and External Progressive Learning (EPL). PL-Net offers the following advantages: 1) IPL divides feature extraction into two steps, allowing for the mixing of different size receptive fields and capturing semantic information from coarse to fine granularity without introducing additional parameters; 2) EPL divides the training process into two stages to optimize parameters and facilitate the fusion of coarse-grained information in the first stage and fine-grained information in the second stage. We conducted comprehensive evaluations of our proposed method on five medical image segmentation datasets, and the experimental results demonstrate that PL-Net achieves competitive segmentation performance. It is worth noting that PL-Net does not introduce any additional learnable parameters compared to other U-Net variants.